Report on the IMKR Colloquium “Digitalization of the Law”

On November 13, 2020, the IMKR hosted a colloquium on the “Digitalization of the Law.” Following opening remarks and an overview of the issues by Prof. Rösler, five legal scholars from Germany, the Netherlands, and the United Kingdom shared their perspectives on current issues related to digitalization.

Prof. Dr. Maximilian Becker, holder of the Chair of Data Protection Law and IT Law at Leibniz University Hannover, kicked off the event with his presentation on “Substantive Justice in Data Protection Law.” Using a diagram, Becker explained that when a user accepts cookies on the website of a major German daily newspaper, 188 servers connect to the user. Most of these are located in the U.S., and over 170 servers are irrelevant to the actual service provided. The servers link data from numerous websites and create a comprehensive profile of the user’s interests, which enables predictions about future behavior. The conflict between data protection and economic interests is addressed in particular by the prohibition on tying in Art. 7(4) of the GDPR. It is intended to limit the unnecessary collection of personal data in the exchange of services. Nevertheless, the legislature’s wording remains deliberately vague. While tying arrangements are problematic due to the risk of market failure, potential lack of transparency, and conceivable situations where users are pressured to disclose data in exchange for services, they are not fundamentally prohibited. However, fair tying practices that avoid pressure and provide transparency are preferable. In this regard, Becker viewed fair freemium models as a middle ground. Here, service providers offer users a choice between free access in exchange for expanded data collection or, alternatively, a paid option that does not involve expanded data collection. In conclusion, Becker discussed the prohibition under Art. 6(1)(a) of the GDPR, as reaffirmed by the Federal Court of Justice (BGH) (NJW 2020, 2540 – Cookie-Einwilligung II). According to this ruling, preset consents to the collection of data that go beyond what is technically necessary are impermissible.

Prof. Dr. Christian Gomille, holder of the Chair of Civil Law, European and Private International Law, and Civil Procedure at Saarland University, then spoke on the topic “The Shitstorm and the Data Subject’s General Right of Personality.” Gomille addressed the discussion of whether a person affected by a “shitstorm” in online media is entitled to protective measures similar to those applicable in the case of a personal siege. In this context, Gomille defined a “shitstorm” as an online phenomenon involving sudden and massive criticism that could not be achieved to the same extent without digital media, since the perpetrators typically reinforce one another. The balancing of interests between those making the statements and the affected individual is complex. The right to freedom of expression and information under Art. 5(1) of the Basic Law (GG), Art. 11 of the Charter of Fundamental Rights of the European Union (CFR), and Art. 10 of the European Convention on Human Rights (ECHR) is counterbalanced, in particular, by the affected person’s right of personality under Art. 2(1) in conjunction with Art. 1(1) GG, Art. 7 CFR, and Art. 8 ECHR. The gradations of the “spheres of privacy” theory provide important guidance for the balancing test to be conducted on the basis of the specific individual case. Ultimately, however, a truly clear-cut legal violation can likely be assumed only in the case of untrue factual assertions. Gomille considers the public shaming effect – which infringes on personality rights – to be present at least when the scope of public exposure exceeds what is justified by the individual’s initial conduct. However, public shaming effects tend to be permissible if the affected individual had previously addressed the public.

In the third presentation, Priv.-Doz. Dr. Jan Oster, LL.M. (Berkeley), Assistant Professor of EU Law and European Institutions at Leiden University (Netherlands), discussed the question: “Is the digitalization of the law possible?” According to current understanding, such digitalization is possible through the use of an “agent” in the form of software that initiates legal processes or makes decisions, such as in the form of legal tech and legal robots. In this context, Oster identifies the problem that software, by strictly following algorithms, functions merely as a mechanism, whereas the law is far more far-reaching and open to development. Furthermore, an algorithm can hardly be reconciled with the requirement to weigh interests, particularly in the area of fundamental rights. Rather, the digitalization of the law is only possible if an autonomous system – as strong artificial intelligence (AI) – possesses the necessary self-learning capabilities. However, as long as digital agents lack genuine knowledge and legal understanding, they can digitalize the law automatically but not autonomously. Consequently, only weak artificial intelligence exists to date. In this context, Oster highlighted the highly controversial question of whether strong AI is even desirable or possible, and when it might be applied.

Priv.-Doz. Dr. Matthias Wendland, LL.M. (Harvard), from the Institute for International Law at Ludwig Maximilian University of Munich, spoke on “Regulatory Strategies to Safeguard Pluralism for Media Intermediaries: Algorithms, Smart Big Data, and Artificial Intelligence as Disruptive Technologies in the Platform Economy.” Wendland began by demonstrating how Google, Twitter, Instagram, and Facebook influence public opinion, resulting in unconscious behavioral change (nudging). This influence is exerted by placing search results and news at the top of the list, as is the case with the Google search engine. As a result, other published information may go unnoticed. Wendland also addressed algorithmic gatekeeping and algorithmic information filtering, which transform quantitative information diversity into qualitative diversity of use and thereby also filter it. Which algorithms lead to a particular ranking remains the central trade secret of media intermediaries. Algorithmic regulation of algorithms offers a way to limit the power of algorithms. For example, an “algorithm certification body” could examine algorithms for discriminatory practices and effects that restrict pluralism. It is also conceivable to require the labeling of algorithms that pose a high risk. This would be based on recurring risk analyses or a real-time evaluation of the number of selected sources in the form of a plurality index with external control mechanisms.

Last but not least, Prof. Dr. Guido Westkamp, LL.M. (London), holder of the Chair of Intellectual Property and Comparative Law at Queen Mary University of London, discussed “Service Provider Liability for Copyright Infringements and the Position of Rights Holders”. The starting point was Directive 2019/790 on copyright and related rights in the Digital Single Market and amending Directives 96/9/EC and 2001/29/EC (commonly known as the Directive on Copyright in the Digital Single Market, DSMD). According to Westkamp, the directive tends to favor rights holders and exploiters over platform operators and users. For example, Article 17 of the DSMD increases platform liability, contrary to the previously applicable host provider privilege under Article 14 of the E-Commerce Directive 2000/31/EC. In particular, a claim for damages will no longer depend on the operator’s knowledge of infringements. However, the threat of a claim for damages could hinder innovation by platforms, which could reduce competitive pressure among rights exploiters and lead to monopolization. Germany intends to address this with the proposed Act on the Adaptation of Copyright Law to the Requirements of the Digital Single Market (UrhDaG-E of September 2, 2020, in German). Thus, content must be collectively licensed by platforms entering into contractual agreements with rights holders (§ 4 UrhDaG-E). Furthermore, Westkamp discussed the user’s ability to flag an uploaded work as a use permitted by contract or law (§ 8 UrhDaG-E). Only if the permission is clearly incorrect must the service provider block the work despite the user’s designation (§ 12 UrhDaG-E). A de minimis threshold for small portions of works would also be introduced (§ 6 UrhDaG-E), as well as an internal complaint procedure (§ 14 UrhDaG-E). Another new feature in German law would be a direct claim by authors against the platform (§ 7 Art. 1 UrhDaG-E), which is likely to significantly curtail the interests of exploiters. This can be justified by the fact that exploiters, unlike authors with their legitimate core interests, do not have their own neighboring rights. However, Westkamp anticipates numerous lawsuits by collecting societies regarding the de minimis threshold. After all, according to the “Infopaq” decision (ECJ, judgment of July 16, 2009 – Case C-5/08, NJW 2010, 753), even small portions may be substantial and thus worthy of protection.